Prompted by the 2023 banking turmoil, policymakers are increasingly working to reform and destigmatize the Federal Reserve’s discount window to bolster banks’ access to liquidity and reduce systemic risk. While there is widespread agreement that banks’ operational readiness to use the discount window benefits them and the broader financial system,[1] there is limited public discourse on how liquidity requirements should reflect this enhanced liquidity position.
Prior BPI blog posts[2] have discussed the need for the banking agencies to recognize and incorporate banks’ ability to access the discount window into regulatory and supervisory liquidity requirements. In this post, we suggest how the liquidity value of discount window capacity could be accounted for in a way that would encourage operational readiness, free up trapped liquidity for some institutions and mitigate the long-standing discount window stigma.
Background
As we have previously written,[3] stigma continues to impede the effective functioning of the discount window. Banks are often hesitant to access the discount window, even when it would be prudent to do so, because of concerns that such borrowing could be perceived as a sign of weakness by market participants and examiners. In its most extreme form, this reluctance has sometimes discouraged banks from even pre-positioning collateral or otherwise being operationally prepared to use the window. However, following renewed focus on the risks of being unprepared or unable to use the discount window after the events of March 2023, a significant number of banks have increased their focus on discount window preparedness. It is important that this work be encouraged; normalizing the use of the discount window is essential to ensuring it serves its intended role as a reliable and effective source of liquidity, particularly during periods of financial stress.
Notwithstanding the importance of the discount window as a liquidity resource, banks are not permitted to factor in their ability to access the discount window for purposes of their liquidity requirements, with few exceptions. The liquidity coverage ratio (LCR), for example, intentionally ignores banks’ pre-pledged capacity at the discount window when considering the composition of a bank’s stock of high-quality liquid assets or its net cash outflows. Internal liquidity stress tests (ILST) similarly do not allow banks to consider the availability of discount window funding, at least for a time horizon of 30 days or less, and there is confusion over the permissibility of its inclusion beyond that time horizon as well.[4] In resolution liquidity requirements, the resolution liquidity adequacy positioning (RLAP) requirement largely mirrors the ILST requirement, and therefore discount window capacity is excluded. For the resolution liquidity execution need (RLEN) requirement, the Federal Reserve and FDIC’s 2019 resolution planning guidance for GSIBs[5] permits banks to assume that discount window funding is available for only a “few days after the point of failure,” which prevents banks from counting on discount window access even after a subsidiary IDI has been recapitalized.
To address these shortcomings, an optimal solution for incorporating a bank’s ability to use the discount window into liquidity planning would achieve three related goals: it would provide an accurate measure of available liquidity, it would encourage operational readiness and it would decrease stigma. This blog post explores a targeted adjustment to the LCR calculation as one possible way to achieve these goals.[6]
Suggested Approach: Adjusted LCR Calculation
Under the LCR framework, a bank’s LCR equals its High-Quality Liquid Assets (HQLA) divided by its projected Net Cash Outflows (NCO) over a 30-day time horizon. The measure is intended to ensure that a bank has sufficient liquidity to meet its anticipated needs under stress at the end of the 30-day window. In the U.S., the projected Net Cash Outflows includes a maturity-mismatch add-on which further requires banking organizations to maintain additional HQLA to cover the peak need within the LCR’s 30-day time horizon.
Access to funding via the discount window could be included in the LCR either by increasing the pool of HQLA available to a bank or reducing its projected Net Cash Outflows. Allowing an institution to reduce its projected Net Cash Outflows would create a strong incentive for banks to preposition collateral at the discount window, including assets that would not typically meet the definition of HQLA, thereby improving the firm’s liquidity practices.
Step 1: Accounting for discount window capacity
An accurate account of discount window capacity should be based upon the unused, pre-pledged non-HQLA collateral a bank has positioned at its regional Reserve Bank, fully accounting for any required collateral haircuts. Including only existing pre-pledged collateral eliminates any potential frictions associated with moving collateral to the discount window (including from the Federal Home Loan Banks). Similarly, accounting for the borrowing capacity on a post-haircut (as opposed to full market value) basis further ensures an accurate calculation of a bank’s actual borrowing capacity. For purposes of the remainder of the post, we will call this amount of borrowing capacity, based on only already-pledged collateral and calculated on a post-haircut basis, a bank’s “Contingent DW Capacity.”
Including Contingent DW Capacity in a bank’s LCR and other liquidity metrics would incentivize banks to be operationally prepared to borrow from the discount window. However, to address the concern that having pre-pledged collateral may not sufficiently ensure a bank can and will actually access the facility during a time of need in the face of possible stigma, we introduce an additional constraint that requires a demonstrable willingness to actually borrow at the discount window, which we will refer to as “Proven Contingent Capacity.” Proven Contingent Capacity is defined as the lesser of 1) the amount a bank has borrowed from the discount window, of any tenor, during the prior calendar quarter and 2) the institution’s current Contingent DW Capacity. Requiring banks to actually borrow from the discount window before it is assumed that they will do so under stress ensures that banks are operationally ready and willing to access the window if needed by requiring them to test and exercise that capability on a quarterly basis. Using the “lesser of” construct also incentivizes banks to maintain their collateral pledges at the discount window.
Step 2: Reducing NCO
Based on this framework, we propose that a bank be permitted to reduce its net cash outflows by its Proven Contingent Capacity when calculating its LCR. This would allow a bank to include, as an offset to its cash outflows, any amount they borrowed from the discount window over the last calendar quarter, so long as it continues to have at least as much available pre-pledged borrowing capacity (as measured on a post-haircut basis). If an institution’s current discount window borrowing capacity (Contingent DW Capacity) is less than the amount borrowed during the prior calendar quarter, then only that Contingent DW Capacity amount would be available to offset outflows. We will refer to a bank’s Net Cash Outflows after this adjustment for Proven Contingent Capacity as the “Adjusted NCO.”
Potential Benefits
Reducing stigma
The approach suggested above would not only create a strong and appropriate incentive for banks to maintain operational readiness but also could start to reduce some of the longstanding stigma associated with using the discount window. A bank would be required to demonstrate its Proven Contingent Capacity quarterly, and therefore any bank wishing to gain the attendant benefit in liquidity metrics would need to access the window on a regular basis. Currently, some institutions undertake relatively small test draws from the discount window to confirm they are able as an operational matter to actually borrow from it, but these test draws do little to combat the stigma. Under our suggested approach, the number of institutions borrowing from the discount window would likely increase as would the overall level of size these borrowings. While stigma is not an easy problem to solve, more institutions borrowing larger amounts as part of their ordinary course of business could chip away at this longstanding problem.
A further potential benefit of this proposal is that an institution potentially identified as borrowing from the discount window through the Federal Reserve’s weekly disclosures would not be assumed to be a troubled institution. Many banks subject to liquidity regulations would be likely to borrow more regularly, and so the appearance of any individual institution borrowing from the discount window would be unlikely to raise substantial public concern. If an institution were to undergo a liquidity stress event, this could essentially provide extra time to deal with the stress as only the continued identification of a bank utilizing the discount window over multiple weeks would draw public scrutiny. This would provide both the bank and examiners with additional time to deal with a potential idiosyncratic liquidity concern.
Releasing trapped liquidity
An additional potential benefit of this approach is that, for some banks, it could also alleviate some concerns related to trapped liquidity at an IDI subsidiary.[7] Under the LCR calculation, a BHC is largely limited to including a subsidiary’s HQLA in the consolidated LCR calculation up to the amount of the subsidiary’s Net Cash Outflows. As a result, a significant amount of firms’ liquidity is “trapped” at the subsidiary IDI and cannot be counted as part of the consolidated organization’s HQLA.[8] Under our suggested construct, a BHC could choose to limit HQLA held at an IDI subsidiary, such that the IDI’s HQLA could be less than its Net Cash Outflows even if such IDI is separately subject to an LCR requirement. This would be possible because the subsidiary could still be compliant with the LCR when factoring in discount window capacity (i.e., its HQLA is greater than its Adjusted NCO). This would allow for less on-balance sheet liquidity at the subsidiary (and therefore less trapped liquidity), freeing up banks’ balance sheets to produce additional credit intermediation for the economy, while ensuring the IDI subsidiary is prepared to use the discount window.
Conclusion
Including a bank’s pre-pledged discount window capacity in liquidity metrics would make such metrics a more accurate of actual liquidity resources and risk. Using Proven Contingent Capacity to reduce NCO under the LCR ensures that banks have demonstrated their ability to access the discount window and continue to maintain collateral at the Reserve Bank to receive benefits under liquidity regulations. This solution would also help reduce discount window stigma, while easing some trapped liquidity issues for certain institutions.
[1] A recent study from Federal Reserve staff found that having pre-pledged collateral with the Federal Reserve can increase the likelihood of a bank borrowing from the discount window and that voluntarily pre-pledging collateral makes the window “a more viable option” in the event of a shock. https://www.federalreserve.gov/econres/notes/feds-notes/pre-pledged-collateral-and-likelihood-of-discount-window-use-20250829.html
[2] https://bpi.com/redundant-costly-and-inaccurate-the-u-s-bank-liquidity-regime-needs-a-rethink/, https://bpi.com/bank-examiners-should-not-be-telling-central-banks-how-to-lend/; https://bpi.com/silvergate-bank-a-discount-window-success-story/; https://bpi.com/clear-recognition-of-the-discount-window-would-improve-liquidity-rules/
[3] https://bpi.com/discount-window-stigma-we-have-met-the-enemy-and-he-is-us/; https://bpi.com/wp-content/uploads/2024/12/BPI-Comment-Letter-RFI-on-Operational-Aspects-of-Federal-Reserve-Bank-Extensions-of-Discount-Window-and-Intraday-Credit-Docket-No.-OP-1838-12-9-2024.pdf
[4] https://bpi.com/bpi-survey-of-bank-treasurers-on-willingness-and-perceived-ability-to-borrow-from-fed-lending-facilities/.
[5] Board of Governors of the Federal Reserve System & Federal Deposit Insurance Corporation, Guidance for 2019 165(d) Resolution Plan Submissions by Domestic Covered Companies that Submitted Resolution Plans in July 2017, 84 Fed. Reg. 1438 (Feb. 4, 2019), available at: https://www.federalregister.gov/documents/2019/02/04/2019-00800/final-guidance-for-the-2019.
[6] While this blog focuses mostly on the LCR, similar adjustments to ILSTs and resolution liquidity requirements, including the removals of prohibitions and impediments for including the discount window, would be needed to fully realize the potential benefits.
[7] https://bpi.com/wp-content/uploads/2021/10/I-Dont-Think-the-LCR-Means-What-You-Think-It-Means-Final.pdf
[8] Banks are also permitted to include additional HQLA “that would be available for transfer to the top-tier Board-regulated institution during times of stress without statutory, regulatory, contractual, or supervisory restrictions, including sections 23A and 23B of the Federal Reserve Act (12 U.S.C. 371c and 12 U.S.C. 371c-1) and Regulation W (12 CFR part 223)”, however practically speaking this does not permit much additional HQLA to be counted.
